Getting accuracy of 94% while doing cosine simmilarity on face encodings

[avirajkhare00]

• face_recognition version: 1.3.0
• Python version: 3.12
• Operating System: OSX

Description

It is given in Readme that modal is having 99% accuracy. While upserting approx 500 face encodings in vector database, we ran the benchmark and found out that accuracy is 94%. We are trying to understand if something went wrong from our end or there is better approach to search across large dataset.
Dataset used: https://www.kaggle.com/datasets/jessicali9530/lfw-dataset